A BOOTSTRAP APPROACH TO EVALUATING THE PERFORMANCEOF AKAIKE INFORMATION CRITERION (AIC) AND BAYESIANINFORMATION CRITERION (BIC) IN SELECTION OF ANASYMMETRIC PRICE RELATIONSHIP

This study addresses the problem of model selection in asymmetricprice transmission models by combining the use of bootstrap methods withinformation theoretic selection criteria. Subsequently, parametric bootstraptechnique is used to select the best model according to Akaike’s InformationCriteria (AIC) and Bayesian Information Criteria (BIC). Bootstrap simulationresults indicated that the performances of AIC and BIC are affected by the size ofthe data, the level of asymmetry and the amount of noise in the model used in theapplication. This study further establishes that the BIC is consistent andoutperforms AIC in selecting the correct asymmetric price relationship when thebootstrap sample size is large.